Post on 19-Aug-2019
DEVELOPMENT OF HEAT FEEDBACK PARAMETER OF
MUSCLE ACTIVATION DETECTION IN FUNCTIONAL
ELECTRICAL STIMULATION
NADHIRAH BTE MOHD KHAIDIR
FACULTY OF ENGINEERING
UNIVERSITY OF MALAYA
KUALA LUMPUR
2012
Univers
ity of
Mala
ya
DEVELOPMENT OF HEAT FEEDBACK PARAMETER OF
MUSCLE ACTIVATION DETECTION IN FUNCTIONAL
ELECTRICAL STIMULATION
NADHIRAH BTE MOHD KHAIDIR
RESEARCH REPORT
SUBMITTED
IN PARTIAL FULFILMENT OF THE REQUIREMENT FOR THE
DEGREE OF MASTER OF ENGINEERING (BIOMEDICAL)
2012
Univers
ity of
Mala
ya
UNIVERSITI MALAYA
ORIGINAL LITERARY WORK DECLARATION
Name of Candidate: NADHIRAH BTE MOHD KHAIDIR
Registration/Matric No: KGL 100012
Name of Degree: Master of Engineering (Biomedical)
Title of Project Paper/Research Report/Dissertation/Thesis (“this Work”):
DEVELOPMENT OF HEAT FEEDBACK PARAMETER OF MUSCLE
ACTIVATION DETECTION IN FUNCTIONAL ELECTRICAL STIMULATION
Field of Study: REHABILITATION ENGINEERING
I do solemnly and sincerely declare that:
(1) I am the sole author/writer of this Work;
(2) This Work is original;
(3) Any use of any work in which copyright exists was done by way of fair dealing and
for permitted purposes and any excerpt or extract from, or reference to or
reproduction of any copyright work has been disclosed expressly and sufficiently
and the title of the Work and its authorship have been acknowledged in this Work;
(4) I do not have any actual knowledge nor do I ought reasonably to know that the
making of this work constitutes an infringement of any copyright work;
(5) I hereby assign all and every rights in the copyright to this Work to the University
of Malaya (“UM”), who henceforth shall be owner of the copyright in this Work
and that any reproduction or use in any form or by any means whatsoever is
prohibited without the written consent of UM having been first had and obtained;
(6) I am fully aware that if in the course of making this Work I have infringed any
copyright whether intentionally or otherwise, I may be subject to legal action or any
other action as may be determined by UM.
Candidate’s Signature Date
Subscribed and solemnly declared before,
Witness’s Signature Date
Name:
Designation:
Univers
ity of
Malaya
ii
ABSTRACT
Functional Electrical Stimulation (FES) was famously being applied in controlling
stimulation of muscles for the restoration of movements in paralysed human. Thus, the
development of sensory feedback is necessary to improve the performance of FES systems
by detecting the level of the muscle force. This research project focuses on the development
of a heat feedback parameter in FES. There are two components for this project. Firstly, the
FES stimulator was constructed to produce a suitable stimulation by modulating the input
parameter. A temperature sensor was used to detect the heat produced by the stimulated
muscle. Secondly, two tests were conducted in several able-bodied subjects. First was
voluntary training and second was using stimulation. During tests, the cycling resistance
and amplitude of pulses current was increased. Then, the heat produced was compared to
the muscle contraction level during the contraction of the vastus lateralis muscles and
determine if either the heat produce is correlated to the level contraction of muscle.
Univers
ityof
Malaya
iii
ABSTRAK
Perfungsian Elektrik Stimulasi (FES) telah terkenal diaplikasikan di dalam
pengawalan rangsangan otot untuk memulihkan pergerakan manusia yang lumpuh. Oleh
itu, pembangunan tindak balas deria adalah diperlukan untuk memperbaiki prestasi sistem
FES dengan mengesan tahap pengaktifan otot. Projek penyelidikan ini memberi fokus
terhadap pembangunan parameter tindak balas haba di dalam FES. Terdapat dua komponen
untuk projek ini. Pertama, sistem perangsang FES telah dibina untuk menghasilkan
rangsangan yang sesuai dengan mengubah nilai parameter input. Alat pengesan suhu telah
digunakan untuk mengesan haba yang dihasilkan oleh otot yang dirangsang. Kedua, dua
eksperimen telah dijalankan terhadap beberapa orang yang sihat serta berupaya.
Eksperimen pertama merupakan latihan secara sukarela dan eksperimen kedua akan
menggunakan rangsangan. Semasa ujian dijalankan, rintangan kayuhan basikal dan
amplitud arus rangsangan telah ditingkatkan. Kemudian, haba yang dihasilkan oleh otot
terangsang akan dibandingkan dengan tahap pengaktifan otot ketika perangsangan otot
’vastus lateralis’ dan menentukan sama ada terdapat perhubungan di antara haba terhasil
dengan tahap pengaktifan otot.
Univers
ity of
Mala
ya
iv
ACKNOWLEDGEMENTS
Assalamualaikum w.b.t.
First of all, I would like to praise to Allah for giving me an opportunity to finish my
research project report titled Development of Heat Feedback Parameter for Muscle
Activation Detection in Functional Electrical Stimulation (FES) by the given duration.
I really would like to thank to my supervisor, Dr. Nur Azah Hamzaid who helped,
advised, stimulating suggestions and encouragement throughout all the time of the project
for writing of this report.
Most of all, I really would like to thank both my parents for supporting and
understand me. Without their guidance, I would not have reach this far. I really appreciate
for all their hard work.
Lastly, for my beloved friend, Nazurah Ahmad Termimi, thank you for all her help,
interest and valuable hint during the constructing the stimulator circuit, testing the
temperature sensors and writing the report.
Thank you once again.
Univers
ity of
Mala
ya
v
TABLE OF CONTENTS
PREFACE
ORIGINAL LITERARY WORK DECLARATION
ABSTRACT……………………………………………………………………………...ii
ABSTRAK………………………………………………………………………………..iii
ACKNOWLEDGEMENTS ……………………………………………………………..iv
TABLE OF CONTENTS………………………………………………………………..v
LIST OF FIGURES……………………………………………………………………...ix
LIST OF TABLES……………………………………………………………………….xii
LIST OF SYMBOLS…………………………………………………………………….xiii
LIST OF ABBREVIATIONS…………………………………………………………...xiv
LIST OF APPENDICES………………………………………………………………...xvi
CHAPTER 1: INTRODUCTION
1.1 Problem Statement…………………………………………………………………..1
1.2 Objectives of the study………………………………………………………………2
1.3 Hypothesis…………………………………………………………………………...2
1.4 Scope of the study…………………………………………………………………...3
1.5 Significance of the study…………………………………………………………….3
Univers
ity of
Mala
ya
vi
CHAPTER 2: LITERATURE REVIEW
2.1 Functional Electrical Stimulation (FES)…………………………………………….4
2.2 Control parameters of electrical stimulation………………………………………...5
2.3 Factors that affect the stimulation…………………………………………………..6
2.4 Physiological effects during electrical stimulation………………………………….7
2.5 Muscle fatigue……………………………………………………………………….8
2.6 Force-fatiguability Relationship……………………………………………………..9
2.7 Circuit of FES stimulator…………………………………………………………..10
2.8 Feedback sensor of FES system……………………………………………………11
2.8.1 EMG………………………………………………………………………..12
2.8.2. Detection of gait phase & monitoring joint angle sensor…………………..13
2.8.3 Pressure sensor……………………………………………………………..14
2.8.4 Fabric stretch sensor………………………………………………………..16
2.9 Relationship between skin temperature and stimulation current…………………..18
CHAPTER 3: METHODOLOGY
3.1 Project Flow Chart…………………………………………………………………20
3.2 First stage: FES stimulator and temperature sensor………………………………..22
3.2.1 FES stimulator……………………………………………………………...22
(a) Part 1: Timer circuit………………………………………………..23
(b) Part 2: Transistors and other electronic devices…………………...24
(c) Part 3: Stimulation electrode………………………………………28
3.2.2 Temperature Sensor………………………………………………………..29
(a) TMP36 temperature sensor………………………………………...29
Univers
ity of
Mala
ya
vii
(b) Arduino Duemilanove……………………………………………...30
(c) Connection between TMP36 and Arduino Duemilanove………….31
3.3 Second part: Conducting Test……………………………………………………...33
3.3.1 Subjects…………………………………………………………………….34
3.3.2 Test 1: Voluntary training………………………………………………….35
(a) Instrument………………………………………………………….35
(b) Placement of temperature sensors………………………………….36
(c) Protocols for Test 1………………………………………………...37
3.3.3 Test 2: FES stimulation…………………………………………………….37
(a) FES Stimulator……………………………………………………..37
(b) Placement of electrode and temperature sensor……………………38
(c) Protocols for Test 2………………………………………………...40
3.4 Third part: Data analysis…………………………………………………………...41
CHAPTER 4: RESULTS
4.1 Output of FES stimulator………………………………………………………….42
4.2 Result of Test 1…………………………………………………………………….45
4.3 Result of Test 2…………………………………………………………………….50
CHAPTER 5: DISCUSSION 56
CHAPTER 6: CONCLUSION
6.1 Summary…………………………………………………………………………...61
6.2 Recommendation for Future Work………………………………………………...62
Univers
ity of
Mala
ya
viii
APPENDICES…………………………………………………………………………….63
BIBLIOGRAPHY………………………………………………………………………...73
Univers
ity of
Mala
ya
ix
LIST OF FIGURES
Figures No.
2.1 FES circuit. (Adapted from Cheng et al., 2004)…………………………………...11
2.2 Placement of FSRs on the surface of subject’s arm.
(Adapted from Amft et al., 2006)…………………………………………………..15
2.3 Block diagram of pressure sensor FES system.
(Adapted from Amft et al., 2006)…………………………………………………..16
2.4 Placement of Fabric Stretch Sensor. (Adapted from Amft et al., 2006)…………...17
2.5 Fabric stretch sensor interface. (Adapted from Amft et al., 2006)…………………17
2.6 Relationship between stimulation current and skin temperature.
(Adapted from Petrofsky et al., 2008)……………………………………………...18
3.1 Project’s flow chart………………………………………………………………...21
3.2 FES stimulator circuit. (Adapted from
http://www.diy-electronic-projects.com/p231-Muscular-Bio-Stimulator)...............22
3.3 Top view of IC LM555 timer.
(Adapted from National Semiconductor Datasheet, LM555 timer)………………..23
3.4 Output Pulse Wave of LM555 timer……………………………………………….24
3.5 Push-button application of AD5220………………………………………………..25
3.6 Output pulse wave………………………………………………………………….26
3.7 Maximum output of pulse current………………………………………………….26
3.8 FES stimulator circuit on a breadboard…………………………………………….27
3.9 Stimulator circuit connected to oscilloscope……………………………………….28
3.10 Stimulation electrodes……………………………………………………………...28
Univers
ity of
Mala
ya
x
3.11 TMP36 temperature sensor.
(Adapted from http://www.ladyada.net/learn/sensors/tmp36.html).........................29
3.12 Output voltage vs. Temperature. (Adapted from Analog Devices
Datasheet, Low Voltage Temperature Sensors, TMP35/TMP36/TMP37)………...30
3.13 Arduino Board Duemilanove. (Adapted from www.arduino.cc).............................31
3.14 Connection between TMP36 and Arduino board.
(Adapted from http://www.ladyada.net/learn/sensors/tmp36.html).........................32
3.15 Five temperature sensors and Arduino board……………………………………...32
3.16 Display of temperature reading…………………………………………………….33
3.17 Aerobike 75XL II…………………………………………………………………..35
3.18 Screen display of Aerobike 75XL II……………………………………………….36
3.19 Placement of temperature sensor on the surface of Vastus Lateralis muscle………36
3.20 FES stimulator used for Test 2. Adapted from (How, 2011)………………………38
3.21 Placement of stimulation electrode………………………………………………...39
3.22 Quadriceps muscles. (Adapted from
http://www.fitstep.com/Advanced/Anatomy/Quadriceps.htm)................................39
3.23 Placement of electrode and temperature sensors…………………………………..40
3.24 Sitting posture for test’s subject. (Adapted from
http://www.shopcompex.com/training/electrode-placements/quadriceps)...............41
4.1 Maximum value of voltage amplitude……………………………………………..43
4.2 Minimum value of voltage amplitude……………………………………………...43
4.3 Minimum value of current amplitude……………………………………………...44
4.4 Maximum value of current amplitude……………………………………………..44
4.5 Linear graph of cycling resistance versus time of cycling…………………………45
4.6 Graph of temperature versus resistance of subject 1……………………………….47
4.7 Graph of temperature versus resistance of subject 2……………………………….49
Univers
ity of
Mala
ya
xi
4.8 Graph of mean and standard deviation of skin temperature
for subject 1 and subject 2………………………………………………………….50
4.9 Graph of temperature versus amplitude current of subject 3………………………52
4.10 Graph of temperature versus amplitude current of subject 4………………………54
4.11 Graph of mean and standard deviation of skin temperature
for subject 3 and subject 4………………………………………………………….55
Univers
ity of
Mala
ya
xii
LIST OF TABLES
Tables No.
3.1 Subject’s personal data……………………………………………………………..34
4.1 Temperature recorded of subject 1…………………………………………………46
4.2 Temperature recorded of subject 2…………………………………………………48
4.3 Temperature recorded of subject 3…………………………………………………51
4.4 Temperature recorded of subject 4…………………………………………………53
Univers
ity of
Mala
ya
xiii
LIST OF SYMBOLS
V Voltage
Ω Ohm resistance
mA miliampere
ms milisecond
Hz Hertz for frequency
F Farad for capacitance
oC Degree Celsius
min. Minutes
R Resistors
W Watt
Univers
ity of
Mala
ya
xiv
LIST OF ABBREVIATIONS
FES Functional Electrical Stimulation
SCI Spinal cord injury
EMG Electromyography
IPI Inter-pulse interval
MP Motor Point
CNS Central nervous system
ATP Adenosine triphosphate
IC Integrated Circuit
OP Operational Amplifier
RMS Root mean squared
MF Median frequency
MAV Mean absolute value
sEMG Surface EMG
pEMG Percutaneous EMG
INT Intervals
Univers
ity of
Mala
ya
xv
PDS Power Density Spectrum
FSRs Force sensitive resistor sensor
LED Light Emitted Diode
ADC Analog-Digital Converter
Eq. Equation
RPM Round per minute
Univers
ity of
Mala
ya
xvi
LIST OF APPENDICES
APPENDIX A Arduino programming for TMP36
APPENDIX B Datasheet of LM555 timer
APPENDIX C Datasheet of AD5220 Digital Potentiometer
Univers
ity of
Mala
ya
1
CHAPTER 1: INTRODUCTION
1.1 PROBLEM STATEMENT
Nowadays, Functional Electrical Stimulation (FES) was famously being applied in
controlling stimulation of muscles for the restoration of movements in paralysed human.
Spinal cord injury (SCI) is one of the troubling injuries that can cause paralyses to the other
part of the human body. Although it has been mentioned that FES can gives significant
medical and physiological benefits towards SCI patient, it may still cause muscle fatigue.
Thus, the development of sensory feedback is necessary to improve the performance of
FES systems by detecting the level of the muscle contraction and thus the forced produced.
Many types of sensory feedback have been developed, but according to the findings, they
are still in a stage of facing some limitations in order to detect the muscle fatigue.
In this Research Project, a new alternative of sensory feedback is developed to
detect the muscle activation by using heat as the output parameter. The relationship
between the muscle contraction and the heat produced by the muscle is examined.
Univers
ity of
Mala
ya
2
1.2 OBJECTIVES OF THE STUDY
The purpose of this project is to develop a feedback parameter of FES to detect
muscle activation using heat as the output parameter. The current sensory feedbacks of the
FES systems still have their own limitations in detecting muscle fatigue.
Therefore, the main objectives of this project are:
i. To reconstruct a Functional Electrical Stimulation (FES) stimulator.
ii. To detect muscle contraction level using heat as the feedback parameter.
iii. To examine the relationship between muscle contraction level and heat production by
the muscle by modulating the stimulation intensity i.e. current of the FES stimulator.
1.3 HYPOTHESIS
This research project focused on the development of a heat feedback parameter in
FES. A stimulator and a set of temperature sensors were design to detect the heat produced
by the stimulated muscle. This stimulator was tested on able-bodied subject to examine the
relationship between the contraction level of stimulated muscle and the heat that produced.
It was predicted that when the muscles level increased, the heat produced by the stimulated
muscle would increased and the relationship between both parameter was linearly
proportional to each other.
1.4 SCOPE OF THE STUDY
Univers
ity of
Mala
ya
3
The scope of this project is to design and develop a heat feedback parameter of FES
for detection of muscle activation. My project has two components. First, a FES stimulator
was build to produce pulses current. In order to produce an effective stimulation, significant
amount of input parameter for the stimulator such as pulse width, amplitude of pulse, and
also the surface of the electrode. Temperature sensor was used to detect the heat produced
by the stimulated muscle. Secondly, experiments were conducted in several able-bodied
subjects where the heat produce will be compared to the muscle contraction level during the
stimulation of the muscles. The objective is to determine if either the heat produce is
correlated to the level contraction of muscle.
1.5 SIGNIFICANCE OF THE STUDY
Currently most FES systems used several types of feedback parameter such as
EMG, force sensor and motor. However, from the previous studies, each of them still has
their own limitation in detecting the muscle activation during stimulation. Their area of
application are more to monitoring, thus did not provide a direct feedback of muscle
activation. This project used a temperature sensor to detect the heat produced by the
stimulated muscle. In the experiment, by modulating the stimulation intensity, the
correlation between heat produced and the contraction of muscle will be examined.
Univers
ity of
Mala
ya
4
CHAPTER 2: LITERATURE REVIEW
2.1 Functional Electrical Stimulation (FES)
Functional Electrical Stimulation (FES) is an approach of controlled stimulation of
muscles for the restoration of movements in paralysed human (Grant, 1988; Tepavac et al.,
1997). FES activates innervated but paralysed muscles by using an electronic stimulator to
deliver trains of pulses to neuromuscular structures (Tepavac et al., 1997). FES also used to
build and maintain the skeletal muscle strength as an addition to physical training for sports
(Currier et al., 1983) (Valli et al., 2002).
Muscles composes of individual muscle fibers that will react to electrical pulses
(Peckham, 1987). The muscles will undergo quick contraction and relaxation which is
called muscle twitch as they react to electrical stimulation (Ew et al., 2005). Thus, the
objective of FES is to elicit safely a controlled stimulation towards the muscle groups
(Sabut et al., 2010).
The criteria of a FES stimulator is that it must be an independent device or portable
with low power utilization, light in weight, small, and easily operated by user (Sabut et al.,
2010) (Popovic, 2006). It is also be able to provide range of frequencies, pulse widths, and
pulse current amplitude (intensities). Each of the parameter can be controlled separately
and programmed to give repeatable treatment (Sabut et al., 2010).
Univers
ity of
Mala
ya
5
Usually, most of the FES stimulator was used to assist the functional movement of a
spinal injury patient (Cheng et al., 2004). Spinal-cord-injury (SCI) is one of the most
disturbing injuries that have been occurred among humans. The spinal injury causes
paralyses towards other part of human body. Paraplegic is a term for paralysed in the legs,
occurred in people who have an injury to the lower part of their spine. As for tetraplegia or
quadriplegia, it is a term for paralyses from the neck or arms level onwards (Grant, 1988).
2.2 Control parameters of electrical stimulation
For controlling the muscle stimulation, there are parameters that involved such as
amplitude of pulse current, frequency and pulse width. During FES, the muscle force output
was controlled by modulating the stimulation intensity and frequency (Kesar et al., 2008).
The modulation of pulse amplitude or the pulse-width can increase or decrease the
recruitment of muscle fibers while the modulation of pulse frequency can increase or
decrease the firing rate of motor neurons (Graham et al., 2006).
For an efficacy of stimulation, the best stimulation for most people is 20 pulses per
second. For achieving the efficacy, the pulse width of stimulators is determined at about
300µs and the amplitude of pulse is depending on individual circumstances. The pulses
must be voltage-controlled and the current does not constant. Thus, the voltage pulses with
the amplitudes of 80 to 150 V. As for the stimulation electrodes, currently surface
electrodes are famously used. The electrodes are placed at the motor-end-point where
motor nerve enters the muscle (Grant, 1988).
Univers
ity of
Mala
ya
6
Another study used a portable, battery-operated, programmable and constant current
stimulator. For the current output, they set the maximum value around 50mA. The pulse
width is between10 to 500µs and the inter-pulse interval (IPI) is about 10 to 100ms which is
the frequency is 10 to 100Hz. As for the electrode, they used the sel-adhesive electrode,
larger electrode for cathode and smaller for the anode (Tepavac et al., 1997).
While, other study mentioned that in general, the amplitude of surface stimulation
ranges from 10mA until 80mA, the stimulation frequency is between 20Hz to 4Hz, and the
pulse width is between 50ms to 300ms, which appears continuously at the surface output
(Sabut et al., 2010).
2.3 Factors that affect the stimulation
There have been mentioned by in previous study, the amount of current required
during electrical stimulation was affected by several factors such as tissue impedance, pad
placement, and shape and size of the electrode. Mostly, stimulation electrode was placed at
a point where it is very sensitive towards stimulation which is called Motor Point (MP)
(Forrester et al., 2004). Another study stated that the MP has a great density of sodium
channels, thus has the lowest impedance. Therefore, the closer the electrode towards MP,
the lesser the stimulation current though its nerve (Reichel et al., 2002).
Previous study investigated the relation between body fat and stimulation current. It
seems that the fat thickness affects the impedance and also the current required for
stimulation. Both quadriceps and gastrocnemius muscle have a thick fat layer compared to
Univers
ity of
Mala
ya
7
biceps muscle. Thus, the current required for stimulate the biceps muscle are lower than
both quadriceps and gastrocnemius muscle (Petrofsky et al., 2008).
A study argued that larger electrode increases tolerance of electrical stimulation.
However, larger electrode can cause the unwanted neighboring muscles to stimulate and
deliver not enough current density to get the desired response (Alon et al., 1994). While
other study discussed that for electrode size, there were no statistical differences for
electrical stimulation as the size difference between the electrodes only about 50%
(Forrester et al., 2004).
2.4 Physiological effects during electrical stimulation
During muscle stimulation, it increases the metabolic requirement with higher rates
of inorganic phosphates and higher cell oxygen level, compared to the natural contraction.
This phenomenon is directly interrelated to the intensity of the stimulation (Forrester et al.,
2004; Vanderthommen et al., 2007).
Cardio-respiratory activity is also affected, with a higher oxygen consumption,
ventilation and respiratory exchange ratio associated with concentric contraction of the
quadriceps femoris induced electrically rather than voluntarily during resistance training
(Theurel et al., 2007).
Univers
ity of
Mala
ya
8
2.5 Muscle fatigue
During the restoration of movement function using FES, the force of the stimulated
muscle eventually will reach its limit which is called muscle fatigue. The muscle fatigue
will limit the effectiveness of FES (Graham et al., 2006). As patient of SCI lost their
sensory pathways, they cannot recognize their muscle fatigue during the stimulation
(Winslow et al., 2003). The muscle contraction via electrical stimulation will decline the
muscle force far more rapidly than when voluntarily contraction. Thus, the occurrence of
muscle fatigue is faster during stimulation of FES in SCI patient (Chesler et al., 1997).
Using FES system, pulses were delivered with a constant amplitude, pulse-width,
and frequency that will stimulate contractions in the same motor units all the time, while
natural contractions allow motor units to rest occasionally. These causes stimulated motor
units to be worn out while the other non-stimulated motor units remain completely inactive.
If the system continuously provides the stimulation, there will be no rest for the
overworked motor units. Therefore, the stimulated muscle will fatigue more rapidly than if
they were activated by the natural stimulation of central nervous system (CNS) (Graham et
al., 2006).
During FES, a declination in force production was due to the fatigue of stimulated
muscle. It could be related basically to changes in the factors below (Tepavac et al., 1997):
(a) Neuromuscular propagation;
M-wave consists of the synchronous sum of all muscle fibre action potentials that
are elicited by the electrical stimulation. The changes in M-waves shows that there are
Univers
ity of
Mala
ya
9
changes in neuromuscular propagation between the site of initiation (axons) and the site of
recording (muscle fibres) or a reduction in the excitability of the muscle fibre membrane.
Low force – long duration voluntary contraction gives out a greater M-wave declination
than the high-force contraction.
(b) Excitation- contraction coupling;
Different tasks activate various mechanisms that can cause force reduction. Usually
the force loss cause by the excitation-contraction coupling can be found after long duration
contractions with slow recovery (30-60 min). Recovery of force loss cause by
neuromuscular propagation and metabolic changes are more rapidly with less than 6
minutes.
(c) Metabolic changes
Muscle force could also decline when the rate of supply ATP and metabolites
generated by the contractile activity cannot support the energy supply. The energy cost of
discontinuous stimulation seems to be higher than that of continuous stimulation.
2.6 Force-fatiguability Relationship
The greater the elicited force, the more rapidly the muscle fatigues. There is also
more fatigue when the force-time integral is greater. The rate and amount of force decline
both increases with the frequency of electrical stimulation (both for whole muscles and for
single motor units) (Tepavac et al., 1997).
Univers
ity of
Mala
ya
10
However, the degree in muscle force may be achieved in two ways which are
varying the number of motor units recruitment and by modulating the firing rate. These
were hypothesized by (Graham et al., 2006), that by randomly modulating the frequency,
pulse amplitude, and pulse-width, would vary the resulting firing rate and level of
recruitment of motor units over time.
2.7 Circuit of FES stimulator
Figure 2.1 shows the example of FES circuit from a previous study (Cheng et al.,
2004). The circuit was divided into two parts. The first part consists of two integrated
circuit (IC) timers 555 and some attached components such as resistors, capacitors, and
diodes. The output of the IC2 is a series of pulses. An additional external trigger signal can
be provided at the sensor input. The frequency and pulse width was controlled by adjusting
the potentiometers (R1, RA and RB) and capacitors (C1 and C2). The second part of the
circuit consists of four operational amplifiers (OP1–4), a transistor, and a transformer. OP1
as an error amplifier, OP2 will amplify the signal to drive the transformer; T1, OP3 and
OP4 are the current-feedback network. The amplitude pulse current can be modulated by
potentiometer, R2. For stepping up the output voltage, the transformer was used. By
stepping up the voltage from 9V to 200V, it will also increase the amplitude pulse current
up to 100mA. By including a current feedback loop, it ensured the current amplitude.
Univers
ity of
Mala
ya
11
Figure 2.1: FES circuit. (Adapted from Cheng et al., 2004)
2.8 Feedback sensor of FES system
Sensory feedback is necessary for effective control of FES systems. The
information about the level of muscle force can improve the performance of FES assistive
systems (Tepavac et al., 1997). Without sensory feedback, the users cannot sense the
fatigue state of their paralyzed muscle during stimulation, thus they may not be able to react
before the contractile failure occurs (Chesler et al., 1997).
Most of current FES stimulator applied an open-loop control function. The
advantage is that the feedback output of sensor did not provide the information about the
effect of stimulations on the subject, plus the user need to control the input parameter
Univers
ity of
Mala
ya
12
manually. So, closed-loop system of the stimulator has been proposed. It would remove the
need for time uncontrollable manual adjustment. However, for this application, it is
important to calculate the expected accuracy of the detected sensor in order to integrate this
information into the parameter’s controller design (Sabut et al., 2010).
Nowadays, there are a few common feedback parameters have been used to detect
the activation of muscle. One of the most famous feedback parameter is by using the EMG.
For analyze the muscle force output, the EMG signal was represented by root mean squared
(RMS), median frequency (MF), and mean absolute value (MAV) (Chesler et al., 1997).
Force sensor and motor sensor also is available to be use as feedback parameter, for
example using f-scan at the patient’s sole to examine the force of patient’s leg during
walking using FES stimulator. As for the common feedback parameter, it is the human
perception. When the muscle stimulate, the patient can sense the movement of the arm or
leg, thus it is consider success for applying the current to the muscle nerve.
2.8.1 EMG
Surface EMG (sEMG) or percutaneous EMG (pEMG) was famously used in the
detection of muscle activation by deriving a voltage measured with electrodes (Tepavac et
al., 1997) (Amft et al., 2006). The advantage of using EMG is because it can be obtained
non-invasively and it reflects the contractile activity of the underlying muscle (Chesler et
al., 1997).
However, the extracting of the muscle force from the sEMG is a difficult task
because of the low signal-to-noise ratio, non-linear behaviour of the muscle force versus
Univers
ity of
Mala
ya
13
muscle length, velocity of shortening, activity of other agonist and antagonist muscles, and
other mechanisms which are dependent on the central nervous system (CNS) (Tepavac et
al., 1997).
Modified surface stimulation and EMG detection equipment were designed and
built to minimize this artifact and to permit detection of the electrical signal generated by
the muscle during contraction. Artifact reduction techniques included shorting stimulator
output leads between stimulus pulses and limiting and blanking slew rate in the EMG
processing stage (Chesler et al., 1997).
Study done by Tepavac et al (1997) was to determine which parameter and
processing technique of the sEMG is best suited to generate the warning signal about
muscle fatigue. After preamplifier and filtering the Semg signal, This signal was used to
derive seven different parameters of the sEMG - four in the time and three in the frequency
domain. In the time domain we calculated RMS, MAV value and fully rectified, integrated
sEMG over 10 ms intervals (INT). In the, frequency domain we calculated power density
spectrum (PDS), MNF and MDF. While the force recorded using strain gauge transducer
(Tepavac et al., 1997).
2.8.2. Detection of gait phase & monitoring joint angle sensor
Previous reports have shown there are two common sensors that were applied in
FES system. The first system was use to detect the gait phase such as swing and stance
phase while the other system was use to monitoring the joint angle, commonly in knee
Univers
ity of
Mala
ya
14
flexion angle. For controlling motion in SCI patient, it is necessary to use both sensors for
detecting gait phase and the position of the joint angle.
Thus, a study had proposed a single sensor system and signal processing system that
could provide real time data, such as gait phases and events, and joint kinematics. They
designed sensor system comprising of Analog Devices accelerometer, rate gyroscopes,
piezoelectric strain gauge and magnetic sensors. At the end, the modified sensor system did
surpass the previous systems that only addressed to one aspect in lower limb control of
FES. These tests suggest that a small, clustered, sensor system worn in conjunction with
other components of a FES system can be designed to provide the variables typically used
for feedback (Williamson et al., 2000).
However, these sensors only manage to detect and control the motion of the patient
but not in term of detecting the muscle fatigue. It is still useless to maintain the controlling
motion in SCI patient as muscle fatigue will limit the effectiveness of FES (Winslow et al.,
2003).
2.8.3 Pressure sensor
Pressure sensor can be also called as force sensitive resistor sensor (FSRs). FSRs is
a polymer thick film device that given out a different values of resistance as the force
applied to its active surface. The resistance is inversely propotional to the applied load.
The resistance value is high up to 1MΩ if the sensor is unloaded, whereas will decreases to
several kΩ if the load is applied to the sensor (Amft et al., 2006).
Univers
ity of
Mala
ya
15
Previous study shows that FSRs is capable to be used to monitor individual
muscles. Figure 2.2 shows the placement of FSRs at the lower arm to detect the muscle
activation (Amft et al., 2006).
Figure 2.2: Placement of FSRs on the surface of subject’s arm.
(Adapted from Amft et al., 2006).
An automatic system controlling the correction of foot drop in hemiplegics was
proposed by another study (Sabut et al., 2010). They used a real-time detection of the foot
pressure using piezo-resistive sensors at insole as a close loop controlled FES system for
the correction of foot drop in hemiplegics. The foot pressure signals from patients were
utilized for signal modulation and hence controlling the stimulation amplitude of the FES
system. As in Figure 2.3, the system includes foot insole sensors, amplifier,
microcontrollers, stimulation unit and stimulating electrodes.
Univers
ity of
Mala
ya
16
Figure 2.3: Block diagram of pressure sensor FES system.
(Adapted from Amft et al., 2006).
Their result showed that the direct feedback from the foot pressure sensitive signals
in a feedback FES system provides a real time control of stimulus current amplitude to
correct foot drop during the swing phase of gait. Since, the stimulation being controlled
automatically which leaves the hands and the mind free could be used to perform other
activities (Sabut et al., 2010).
2.8.4 Fabric stretch sensor
Fabric stretch sensor can be used as a strain sensor. When the fabric was stretched,
will resulting changes of the electrical resistance. The resistance increases when the fabric
elongated. However, the sensor can become a problem to some applications due to highly
non-linear of the resistance in its response to strain and cause a hysteresis. Despite the
disadvantages, the sensor still be attempted to be used to detect body postures and arm
gestures (Lorussi et al., 2005).
Univers
ity of
Mala
ya
17
In previous study, the fabric stretch sensor attached to the lower arm in Figure 2.4
was used to capture muscle activations during specific hand and arm activities. For
measuring the changing resistance of the fabric stretch sensor, they use a Wheatstone
bridge depicted in Figure 2.5. However by the end of the study, it shows that fabric stretch
sensor cannot be used to monitor the individual muscles as it is limited due to strong
hysteresis (Amft et al., 2006).
Figure 2.4: Placement of Fabric Stretch Sensor.
(Adapted from Amft et al., 2006).
Figure 2.5: Fabric stretch sensor interface.
(Adapted from Amft et al., 2006).
Univers
ity of
Mala
ya
18
2.9 Relationship between skin temperature and stimulation current
A study had included the investigation of relationship between skin temperature and
stimulation current. Skin temperature was measured with a BioPac skin temperature
thermistor probe. The three conditions which are; normal room exposure, five minute
exposure of cold pack and five minute exposure of hot pack on the surface of the muscles.
Figure 2.6 shows that the skin temperature was different between the three conditions and it
was almost linear. Based on the equation in the figure, the stimulation current was
increased by 0.54 mA for every degree increase in skin temperature (Petrofsky et al., 2008).
Figure 2.6: Relationship between stimulation current and skin temperature.
(Adapted from Petrofsky et al., 2008).
Sim
ula
tio
n c
urr
en
t (m
A)
Univers
ity of
Mala
ya
19
Previous study have shown that by applying a heating towards the skin surface will
shifts the currents in the skin during electrical stimulation. Blood has a lower electrical
resistance. By increasing the blood flow in the skin, the currents were shifted to that area.
When the skin were heated by a hot pack, it increased the blood circulation, thus increased
the current required to stimulate the muscle (Petrofsky et al., 2007).
Previous studies mentioned above, they investigated and recorded the value of
stimulation current as the skin temperature changes through different conditions. As for this
research, a new investigation was take place to study whether the change of stimulation
current will also affect the skin temperature and produce a linear relationship.
Univers
ity of
Mala
ya
20
CHAPTER 3: METHODOLOGY
3.1 Project Flow Chart
This section describes the procedure of constructing; experimenting and analyzing
in order to achieve the main objectives. Flow chart in Figure 3.1 shows the procedures
throughout this study. The whole procedure was divided into three stages. The first stage
involved with the construction of FES stimulator and temperature sensors. For building the
FES stimulator, an integrated circuit (IC) timer 555 was used to produce a pulse wave.
Then, circuit timer was connected to other analog components to produce required current
amplitude for muscle stimulation. Skin surface stimulation electrode was used to transfer
the stimulation current to the specific muscles. For detecting the muscle activation, this
study used a temperature sensor as a feedback sensor to detect changes of skin temperature
during stimulation. An Arduino microcontroller board was used to read the voltage output
of temperature sensor, thus display it in form of degree temperature. After successfully
construct the FES stimulator, at the second stage, two tests were conducted towards the
subjects. Test 1 involved the study of relationship between cycling resistance and skin
temperature. This test was carried out as a voluntary training by subjects. Test 2 involved
the study of relationship between current amplitude and skin temperature. At the third
stage, data analysis was taken place. For both tests, the correlation between the muscle
activation and the muscle heat production was examined.
Univers
ity of
Mala
ya
21
Figure 3.1: Project’s flow chart.
Build FES stimulator
Timer circuit
Surface
electrode
Test 1
Test temperature sensor
Voluntary cycling
Aerobike
Increase the
cycling resistance
Data analysis:
Relation between
muscle activation and
heat production.
Compare data between
Test 1 and Test 2.
Build temperature sensor
TMP36
Arduino
First stage
Second stage
Third stage
Start
End
Test 2
Test temperature sensor
with FES stimulator
Stimulate muscle
Difference current
amplitude
Conduct test
Univers
ity of
Mala
ya
22
3.2 First stage: FES stimulator and temperature sensor
3.2.1 FES stimulator
Before constructing the actual circuit of FES stimulator, Multisim software was
used to construct a non-real circuit and the simulation was done in order to study the
connection of the circuit and produce the required outcome. By going through this step, it
helped by reducing the complication during constructing the actual circuit.
The FES stimulator was comprises of three parts, as shown in Figure 3.2. The first
part consists of IC LM555 timer and a group of resistors and capacitors. The second part
consists of resistors, two PNP transistors and a step-up transformer. Lastly, the third part
was the stimulation electrodes that will deliver the stimulation current to the muscle.
Figure 3.2: FES stimulator circuit.
(Adapted from http://www.diy-electronic-projects.com/p231-Muscular-Bio-Stimulator)
Univers
ity of
Mala
ya
23
(a) Part 1: Timer circuit
The IC LM555 timer circuit in Figure 3.3 was suitable to be used as it is a highly
stable device for generating accurate time delays or oscillation. For astable operation as an
oscillator, the included external components are resistors, capacitors and diodes will
accurately control the free running frequency and duty cycle.
Figure 3.3: Top view of IC LM555 timer.
(Adapted from National Semiconductor Datasheet, LM555 timer)
For setting up the value of frequency pulses and duty cycle, the value of R1, R2 and
C2 was controlled. According to the datasheet for LM555 timer, the formula in Eq. 1 was
used to obtain the frequency pulses. In this project, it had been decided to set the value of
frequency pulses to 80Hz. As in previous studies, most of them set up the range of
frequency pulses between 10 – 100Hz. The calculation below shows how to obtain f =
80Hz and the value for R1 = 4kΩ, R2 = 7kΩ while C2 = 1µF.
f = 1.44 / (R1 + 2R2) C2 ------------- (Eq. 1)
f = 1.44 / [4kΩ +2 (7kΩ)] 1µF ------------- (Eq. 2)
f = 80Hz ---------------- (Eq. 3)
Univers
ity of
Mala
ya
24
For duty cycle, this formula in Eq. 4 was used to obtain approximate 38%. For the
calculation;
D = R2 / (R1 + 2R2) ------------- (Eq. 4)
D = 7kΩ / 4kΩ +2 (7kΩ)] ------------- (Eq. 5)
D = 0.3888 x 100 ------------- (Eq. 6)
D ≈ 38% ------------- (Eq. 7)
Aside the calculation theory, the oscilloscope was connected to the output of
LM555 timer to display the output pulse wave. Figure 3.4 shows the display of the
oscilloscope. It also shows the frequency pulse is 80Hz and the maximum output voltage is
9V.
Figure 3.4: Output Pulse Wave of LM555 timer.
(b) Part 2: Transistors and other electronic devices.
According to the previous Figure 3.2, second part of the stimulator circuit consists
of two PNP Transistor, Q1 and Q2 (2N3906), some resistors, a 10kΩ digital potentiometer
Univers
ity of
Mala
ya
25
(AD5220), and LED1 (green). The output of LM555 timer was connected to Q1. Q1 acts as
a buffer to control the flow of current. While Q2 inverts the pulses polarity.
For the 10kΩ potentiometer, it controls the amplitude of pulse current and
approximately displayed by LED1 brightness. Originally, the circuit uses an analog
potentiometer, but it was replaced with a digital potentiometer (AD5220) as in Figure 3.5,
that provides a push-button application.
Figure 3.5: Push-button application of AD5220.
Potentiometer has three pins which are A1, B1 and W1, the wiper. For this
connection in Figure 3.5, the pin A1 was connected to VCC which is 3V. While pin B1 was
connected to 6.8kΩ as in Figure 3.2. Pin W1 was connected to the transistor PNP, Q2.
The push button application of AD5220 is only a simple application to increment
or decrement the value of resistance. The switch button, J1 in Figure 3.5 determined either
to increment or decrement. As the switch connected to pin U/D of AD5220, the resistance
Univers
ity of
Mala
ya
26
of the potentiometer would increment when the push button was applied repeatedly. It
reduced the brightness of LED1. When the switch connected to the ground, the resistance
would decrement when applied push button. Thus, it increased the brightness of LED1.
The outcome that produced by the stimulator were displayed using an oscilloscope
and a multi-meter. Figure 3.6 show the oscilloscope displayed the pulse wave with the
maximum amplitude voltage 2.77V and the pulse frequency 80Hz.
Figure 3.6: Output pulse wave.
For displaying the value of current amplitude, multi-meter was used. The maximum
amplitude of pulse current produced by the circuit simulation is 48.3mA as shown in Figure
3.7.
Figure 3.7: Maximum output of pulse current.
Univers
ity of
Mala
ya
27
After completing the circuit simulation, the construction of actual circuit was take
place. The construction was done on a breadboard as a temporary circuit as it is easy to
connect parts of the circuit without soldering. The outcome result of the circuit was tested
and recorded first before transferring the circuit permanently onto a circuit board with
soldering.
A 9V battery was used as the power supply for the whole circuit except the digital
potentiometer (AD5220); instead it used two batteries of 1.5V which total of 3V. Figure 3.8
shows the entire stimulator circuit. The red switch, J1 connected between 9V battery and
the whole circuit. It works as an ON/OFF switch for the circuit. The blue switch is a switch
to determine the increment or decrement value of resistance of AD5220. Figure 3.9 show
the circuit was connected to the oscilloscope to display the voltage pulse wave.
Figure 3.8: FES stimulator circuit on a breadboard.
Univers
ity of
Mala
ya
28
Figure 3.9: Stimulator circuit connected to oscilloscope.
(c) Part 3: Stimulation electrode
For the stimulator’s electrode, two self-adhesive 4.5cm x 9cm surface electrodes
(EMPI, STIMCARE) were used as in Figure 3.10. The electrodes were used for a single
patient, non-sterile, reusable and self-adhering. Hydro-gel was applied between the surface
of electrode and surface of the skin as the gel helped to conduct the pulse current during
stimulation.
Figure 3.10: Stimulation electrodes.
Univers
ity of
Mala
ya
29
3.2.2 Temperature Sensor
For the detection of muscle activation, this study proposed to use simple, low cost
sensors to meet the following requirements: unobtrusive integration into garments, small
power consumption for extended lifetime and simple interfacing. The investigation of
temperature sensor to detect the muscle activation was done by sensing the heat changes of
skin surfaces.
(a) TMP36 temperature sensor
For this project, a temperature sensor (TMP36) was used to detect the heat
temperature of the skin surface above the stimulated muscle. In order to get more accurate
data, five temperature sensors were used and placed along the skin surface of the stimulated
muscle. TMP36 is a low voltage, precision centi-grade temperature sensor. It has three pins
which indicate 2.7 – 5.5Vin, analog voltage out and ground as shown in Figure 3.11.
Figure 3.11: TMP36 temperature sensor.
(Adapted from http://www.ladyada.net/learn/sensors/tmp36.html)
Univers
ity of
Mala
ya
30
According to the ‘b’ line in Figure 3.12, TMP36 provides a voltage output which
linearly proportional to the Celsius temperature and it is specified from −40°C to +125°C,
provides a 750 mV output at 25°C.
Figure 3.12: Output voltage vs. temperature.
(Adapted from Analog Devices Datasheet, Low Voltage Temperature Sensors,
TMP35/TMP36/TMP37)
(b) Arduino Duemilanove
The output of the TMP36 is in analog voltage, thus to read the temperature value
from the sensor; need to plug in the output pin directly into an Analog Digital Converter
(ADC) input. In this project, a microcontroller board, Arduino Duemilanove was used to
change the output of voltage and display it in form of temperature value.
As in Figure 3.13, Arduino Duemilanove has six analog inputs, a USB connection,
and a reset button. It contains everything needed to support the microcontroller; simply
Univers
ity of
Mala
ya
31
connect it to a computer with a USB cable or power it with a AC-to-DC adapter or battery
to get started.
Figure 3.13: Arduino Board Duemilanove.
(Adapted from www.arduino.cc)
(c) Connection between TMP36 and Arduino Duemilanove
The connection between TMP36 and the Arduino board was simple. Figure 3.14
shows the connection between the sensor and the board. The Vin pin of sensor will
connected to either pin 3.3V or 5V on the Arduino board. As for this project, pin 3.3V was
used, which mean 3.3V was supplied to the temperature sensor. The leg of analog voltage
output was connected to the pin analog IN ‘0’ on the board. As for the ground leg of sensor,
it was connected to the ground pin on the board.
However, this project was using five temperature sensors which are S1, S2, S3, S4
and S5, thus each of analog voltage output of each temperature sensor were connected to
different analog IN pin; A0, A1, A2, A3, and A4 on the Arduino board. According to
Univers
ity of
Mala
ya
32
Figure 3.15, the output of S1 connected to A0, these connections goes on until S5 that
connected to A4. The Vin and ground of each sensors were arranged in series on a circuit
board, then were connected to pin 3.3V and ground pin on the board.
Figure 3.14: Connection between TMP36 and Arduino board.
(Adapted from http://www.ladyada.net/learn/sensors/tmp36.html)
Figure 3.15: Five temperature sensors and Arduino board.
Univers
ity of
Mala
ya
33
For the temperature reading, the listing command to read all five TMP36 sensors
showed in Appendix A was programmed into the arduino board. Figure 3.16 shows the
display of temperature reading in the computer. C0 until C4 indicate the sensors S1 until
S5. The reading was programmed to update every two seconds.
Figure 3.16: Display of temperature reading.
3.3 Second part: Conducting Test
Second part of this project was divided into two tests. Test 1 involved with
voluntary training to test the accuracy of the temperature sensor and to study the
relationship between the cycling resistance and the skin temperature. Test 2 used the FES
stimulator to stimulate specific muscles. During stimulation, the temperature sensor was
used to detect the changes of skin temperature.
Univers
ity of
Mala
ya
34
3.3.1 Subjects
In this project’s experiment, four individuals were recruited as test subjects. Four of
them are 16 to 24 years old males and are able-bodied individuals. The inclusion criteria for
the study include having a healthy body and normal weight individual. Two subjects were
arranged to do Test 1 and the remaining two subjects were arranged to do Test 2. Table 3.1
shows the personal data of participated subject.
For subject’s preparation before conducting both tests, they were advised not to
have any heavy exercise within five to six hours, have a meal within one to two hours and
wear a pair of short pants for easily placed the stimulation electrode and temperature
sensors on the thigh surface.
Table 3.1: Subject’s personal data.
Age Height (cm) Weight (kg)
Test 1
Subject 1 17 166 66
Subject 2 20 176 74
Test 2
Subject 3 24 167 69
Subject 4 21 170 70
Univers
ity of
Mala
ya
35
3.3.2 Test 1: Voluntary training
(a) Instrument
For Test 1, an Aerobike 75XL II in Figure 3.17 was used by the subjects for
voluntary cycling. The aerobike can be used for sport training. The parameters that can be
controlled by the bike system are the resistance in Watt and the speed of the cycling in
round per minute (rpm). But to control the parameters is up to the capability of the user to
maintain the cycling performance.
Figure 3.17: Aerobike 75XL II.
Figure 3.18 shows the screen display of the aerobike. For maintaining the good
performances, the user must maintain the speed of cycling in range of 46 rpm to 52 rpm. As
the needle on the screen remains in good condition, the resistance of cycling would be
increased by 1 Watt for each 5 seconds. Thus, the longer the time of cycling, the resistance
will increase bit by bit.
Univers
ity of
Mala
ya
36
Figure 3.18: Screen display of Aerobike 75XL II.
(b) Placement of temperature sensors
Cycling activity was involved by a certain group of muscles; one of the group
muscles is the quadriceps muscles. Vastus lateralis of quadriceps muscles was chosen to be
tested for this voluntary training. Five temperature sensors were aligned along the skin
surface of vastus lateralis muscle from distal to proximal as in Figure 3.19. S1 was placed
at the very distal of the muscle and S5 was placed at the very proximal of the muscle.
Figure 3.19: Placement of temperature sensor on the surface of Vastus Lateralis muscle.
S1 S3 S2 S5 S4 Univers
ity of
Mala
ya
37
(c) Protocols for Test 1
The first subject was asked to roll-up only one side of the short pants so that it
would be easy to place the sensors. Then, the subject was ordered to sit on the Aerobike
75XL II to get ready for cycling. The temperature sensors were placed as mentioned
previously and leave them for a few seconds before the initial skin temperature recorded.
Subject’s personal data was taken such as height, weight and age. After done entering the
data into the aerobike’s memory, need to wait for one minute before start the cycling. The
subject was asked to cycle for 20 minutes straight without stop and maintained the good
performance of cycling. During the cycling duration, the resistance of cycling was
increased, thus each 5 minutes; the temperature of each sensor was recorded. All the above
procedures were repeated for the second subject.
3.3.3 Test 2: FES stimulation
(a) FES Stimulator
For Test 2, the constructed stimulator was replaced with a previous stimulator that
was done by an undergraduate student (How, 2011). The problem faced by the constructed
stimulator will be discussed in Chapter 5. Figure 3.20 shows the FES stimulator that was
used for Test 2. For setting up the stimulator, firstly the controlled parameter was set up. In
this study, the amplitude of the pulse was modulated, while the other parameters were set as
constant. The maximum voltage amplitude was ≈7V and the frequency pulse was 17Hz. By
modulating the resistance value of potentiometer, the current output was modulated. The
Univers
ity of
Mala
ya
38
range of current output that modulated was between 0.85mA to 1.55mA which will be
increment for every five minutes. The total time to stimulate the muscle was 20 minutes.
Figure 3.20: FES stimulator used for Test 2.
Adapted from (How, 2011).
(b) Placement of electrode and temperature sensor
The two surface electrodes were placed on the skin surface of the same muscle. For
this project, the electrodes were at the proximal and distal of Vastus Lateralis muscle as in
Figure 3.21. Vastus Lateralis is one of the four quadriceps muscles as shown in Figure 3.22.
Before placing the electrodes on the subject’s surface thigh, the skin was washed and dried
in order to ensure good conductivity and electrodes were placed only on healthy skin. A
light finger pressure was applied to the entire electrode surface. The electrodes were stayed
on the skin better when they reached the skin temperature.
Univers
ity of
Mala
ya
39
Figure 3.21: Placement of stimulation electrode.
Figure 3.22: Quadriceps muscles.
(Adapted from http://www.fitstep.com/Advanced/Anatomy/Quadriceps.htm)
As for the temperature sensors, five of them were placed in series along the skin
surface from proximal to distal as shown in Figure 3.23.
Univers
ity of
Mala
ya
40
Figure 3.23: Placement of electrode and temperature sensors.
(c) Protocols for Test 2
For Test 2, two subjects were participated. The first subject was asked to sit on a
chair with a 90o posture as in Figure 3.24. The stimulation electrode and temperature
sensors were placed as mentioned previously. The personal data of patient; height, weight
and age were taken. Before starting the stimulation, the initial temperature of muscle
surface was read by the arduino. For starting, the output current was modulated until the
subject can feel the presence of current and feel comfortable. The first five minutes, the
temperature was recorded and the current amplitude was increment from 0.85mA to
1.15mA. These steps were repeated for each 5 minutes until reach the maximum value of
current, 1.55mA.The duration of this whole stimulation was 20 minutes. All the procedures
were repeated for the second subject.
S1 S2 S4 S3 S5
Stimulation electrode
Univers
ity of
Mala
ya
41
Figure 3.24: Sitting posture for test’s subject.
(Adapted from http://www.shopcompex.com/training/electrode-placements/quadriceps)
3.4 Third part: Data analysis
The data obtained from Test 1 was analyzed to study the relationship between the
increasing of cycling resistance and the skin temperature. For Test 2, the data obtained to
examine the relationship between stimulation current and the skin temperature. All the data
were list in tables and plotted into graphs using Microsoft Office Excel to examine the
correlation. Then, data between Test 1 and Test 2 were compared to examine the
correlation between cycling resistance and stimulation current by observing the respond of
skin temperature.
Univers
ity of
Mala
ya
42
CHAPTER 4: RESULTS
This chapter discussed the outcome of FES stimulator and the results obtained from
both Test 1 and Test 2. The output of actual stimulator was compared with the output of
simulation stimulator. As for Test 1 and Test 2, both of them showed a relationship
between cycling resistance, stimulation current and skin temperature, thus lead to the
finding of correlation between muscle activation and the heat produced by the muscle.
4.1 Output of FES stimulator
After the FES stimulator was constructed, the output value was tested first before
connected to stimulation electrode and start the stimulation. The value outputs such as pulse
frequency, pulse width and current amplitude were displayed using an oscilloscope and
ammeter. This was to make sure either the value output were the same as the output of
simulation stimulator.
As mention previously in Chapter 3, the 10kΩ digital potentiometer was used to
control the level of current amplitude as well as the voltage amplitude. The value of voltage
amplitude and pulse frequency was displayed using an oscilloscope. Figure 4.1 and Figure
4.2 shows the maximum and minimum of voltage amplitude and the value of pulse
frequency. When the potentiometer was set at the highest value of resistance, the voltage
Univers
ity of
Mala
ya
43
amplitude increased until 9V. As for the opposite, the lowest value of resistance had made
the voltage amplitude to decrease until 1.60V and the pulse frequency for both minimum
and maximum voltage remained the same, which was already set earlier during the circuit
construction.
Figure 4.1: Maximum value of voltage amplitude.
Figure 4.2: Minimum value of voltage amplitude.
Univers
ity of
Mala
ya
44
For this study, the range of current output that supposedly to be tested was between
10mA and 40mA. As for the testing, the potentiometer was set to the lowest resistance, to
display the minimum value of current amplitude. Figure 4.3 shows that the lowest value of
current amplitude obtained was around 3mA while in Figure 4.4, it shows the maximum
value of current amplitude with 79mA.
Figure 4.3: Minimum value of current amplitude.
Figure 4.4: Maximum value of current amplitude.
Univers
ity of
Mala
ya
45
4.2 Result of Test 1
For Test 1, both subjects were asked to voluntary cycled for 20 minutes. The
cycling speed must be maintained in range of 46 – 52 rpm in order to get good
performance. In the experiment, increased value of power indicates the increased of cycling
resistance. While maintaining the performance, the cycling resistance will increase due to
increased of 1 Watt of power for every five seconds. Through calculation, the value of
resistance can be obtained for each time. For example:
1 minute = 60 seconds
10 minutes x 60s = 600 seconds
1 Watt = For every 5 seconds
Thus, Figure 4.5 shows that the cycling duration is linearly proportional to the cycling
resistance.
Figure 4.5: Linear graph of cycling resistance versus time of cycling.
0
50
100
150
200
250
0 5 10 15 20
Po
wer
(Watt
)
Time (Minutes)
Cycling resistance (Power) vs Time
Univers
ity of
Mala
ya
46
Table 4.1 shows the list value of cycling resistance, time of cycling and recorded
temperature from five sensors of subject 1. The temperature recorded were the skin
temperature of the surface skin above vastus lateralis muscles. For all five sensors, the
initial skin temperature was between 28oC and 29
oC. As the subject start cycle, the time and
cycling resistance were examined. At minute 5 value of power increased to 60 Watt,
supposedly the temperature would also increase. However, the temperature value was
dropped into a range of 25oC until 28
oC for all five sensors. Then, after minute10 until
minute 20, the temperature values were increased until reached a range value of 29oC to
31.5oC.
Table 4.1: Temperature recorded of subject 1.
Power
(Watt)
Time Temperature sensor (Degree Celcius)
S1 S2 S3 S4 S5
0 Initial 28.12 28.12 28.61 29.1 29.1
60 Minute 5 27.15 26.66 25.63 28.12 27.64
120 Minute 10 28.22 27.15 27.15 28.61 29.1
180 Minute 15 28.61 29.1 29.59 29.59 30.08
240 Minute 20 29.59 31.01 30.08 30.57 31.54
Graph in Figure 4.6 shows the relationship between cycling resistance and the skin
temperature of subject 1. From the graph, it showed that the skin temperature was increased
as the value of power increased. For the first 5 minute, there might be a problem among the
temperature sensors that lead to the reduction of temperature value recorded. For initial
Univers
ity of
Mala
ya
47
value, temperature recorded by S4 and S5 were the highest while temperature recorded by
S1 and S2 were the lowest. By the end of the test, it showed that starting from S1 until S5,
the temperature recorded was increased. This means that the skin temperature at the distal
of vastus lateralis was the lowest and the proximal part of the muscle had the highest
temperature.
Figure 4.6: Graph of temperature versus power of cycling of subject 1.
Temperature recorded for all five sensors of subject 2 were listed in the Table 4.2.
The initial skin temperature of subject 2 was in range of 24oC to 26
oC. Five minute after
cycling was started; the temperature for each sensor was increased. As the time of cycling
reach 10 minutes, the temperature still increased. However, as it reached 15 minutes of
cycling, temperature value from S2 and S4 still maintained the same and the value from S5
was dropped nearly 1oC. While temperature value of S1 and S3 still increased. End of the
25
26
27
28
29
30
31
32
0 60 120 180 240
Tem
per
atu
re (
Deg
ree
Cel
ciu
s)
Power (Watt)
Temperature versus Power (Subject 1)
S1
S2
S3
S4
S5
Watt
Univers
ity of
Mala
ya
48
test, there was a change in the performance of each sensor. S1 gave out a same temperature
as before, value of S2 was slightly increased, value of S3 was slightly decreased, and value
of S4 and S5 was increased. The mean temperature that recorded by five sensors was about
28.07oC.
Table 4.2: Temperature recorded of subject 2.
Power
(Watt)
Time Temperature sensor (Degree Celcius)
S1 S2 S3 S4 S5
0 Initial 24.71 25.68 24.22 25.68 26.17
60 Minute 5 26.66 28.12 25.68 26.66 27.64
120 Minute 10 27.15 28.12 26.66 28.12 28.12
180 Minute 15 27.64 28.12 27.64 28.12 27.15
240 Minute 20 27.64 28.61 27.15 28.64 28.31
Figure 4.7 shows that the pattern of temperature increased of subject 2 was slightly
different compared to subject 1. The skin temperature was theoretically increased linearly
with the increased of power value of cycling resistance and time of cycling. However, in
the relationship graph of subject 2, the temperature recorded was not constantly increased
as some of the sensors gave out a decreased of temperature during cycling. For the initial
temperature, value of S5 was the highest while value of S3 was the lowest. Yet, at the end
S4 gave out the highest value and S3 remained the lowest.
Univers
ity of
Mala
ya
49
Figure 4.7: Graph of temperature versus resistance of subject 2.
The mean and standard deviation graph in Figure 4.8 shows that pattern of
temperature increased of subject 2 was better than subject 1. For subject 2, the mean
temperature keep increased when power increased. For subject 1, the temperature decreased
during the power value of 60Watt. But after that, start to increased back until the maximum
value of power. Then, the mean difference between end temperature and initial temperature
of subject 1 and subject 2 were different. For subject 1, the mean difference was
approximate to 2oC while for subject 2, the mean difference was nearly 3
oC. It showed that
during the 20 minutes cycling, the rate of increased temperature for subject 2 was slightly
higher than subject 1. The standard deviation among the five sensors for both subject 1 and
subject 2 was different for each value of resistance where the temperature was recorded.
24
25
26
27
28
29
0 60 120 180 240
Tem
per
atu
re (
Deg
ree
Cel
ciu
s)
Power (Watt)
Temperature versus Power (Subject 2)
S1
S2
S3
S4
S5
Watt
Univers
ity of
Mala
ya
50
Figure 4.8: Graph of mean and standard deviation of skin temperature for subject 1 and
subject 2.
4.3 Result of Test 2
For Test 2, both subjects were not asked to do any voluntary training, instead they
were applied with a stimulate current to stimulate their vastus lateralis muscles. As
mentioned before, the stimulator that have been constructed had failed to be applied on
human muscle, thus been replaced with another stimulator to carry on this test.
The new stimulator was set to the maximum value of pulse frequency was 17Hz and
the output voltage pulse was ≈7V. As for the output current, the maximum value was
1.55mA. Thus, the initial skin temperature was taken before started the stimulation.
Table 4.3 shows the recorded temperature from five sensors of subject 3. For all
sensors, the initial skin temperature for subject 3 was between 24oC and 27
oC. For starting,
24
25
26
27
28
29
30
31
32
0 50 100 150 200 250
Mea
n &
Sta
nd
ard
Dev
iati
on
(Deg
ree
Cel
ciu
s)
Power (Watt)
Mean & Standard Deviation versus Power (Subject 1 & 2)
Subject 1
Subject 2
Watt
Univers
ity of
Mala
ya
51
the stimulator was set to give out 0.85mA for about 5 minutes. At the end of 5 minutes,
value of temperature was recorded. Supposedly the temperature was increased for all
sensors, but S5 gave out a slightly decreased value. Then, the output current was increased
to 1.15mA for another 5 minutes. This went on until minute 20 with the maximum current
of 1.55mA. At the end, it showed that temperature for all sensors were increased into a
range of 25oC until 28
oC.
Table 4.3: Temperature recorded of subject 3.
Current
(mA)
Time Temperature sensor (Degree Celcius)
S1 S2 S3 S4 S5
0 Initial 25.68 26.66 26.17 27.12 24.71
0.85 Minute 5 25.68 26.66 26.66 28.12 24.23
1.15 Minute 10 26.66 27.12 27.12 29.01 25.11
1.35 Minute 15 26.02 27.64 27.17 28.64 25.21
1.55 Minute 20 26.17 27.66 27.66 28.62 25.68
Figure 4.9 shows the graph of a relationship between current amplitude and the skin
temperature of subject 3. From the graph, it showed that the skin temperature was increased
as the current amplitude increased. However, the increased pattern of temperature was not
linear with the increased of current amplitude as some of the sensors detect a decreased
value of temperature during the stimulation. For initial temperature value, S4 gave out the
highest value while temperature detected by S5 was the lowest. As the muscles were
stimulated with 1.55mA, seems that S4, S2, and S3 gave out a higher reading temperature
Univers
ity of
Mala
ya
52
compared to S1 and S5. Concluded that both proximal and distal muscles that located near
the electrodes produced lower temperature while the medial muscle which is far from both
electrodes produced a higher temperature.
Figure 4.9: Graph of temperature versus amplitude current of subject 3.
For subject 4, the list of temperature recorded for were listed in the Table 4.4. The
initial skin temperature of subject 4 was in range of 26oC to 28
oC. Five minute after
stimulation started; temperature for three sensors were increased while the other two were
remained the same as the initial reading. When the current increased for another five
minutes, most of temperature was increased except temperature at S5 was decreased. The
current increased and time has reached 15 minutes of stimulation, temperature value from
S3 maintained the same and the others increased. End of stimulation, S1 and S5 gave a
difference of 1oC higher compared to other sensors. Thus, the range of temperature reading
was between 28oC and 29
oC.
23
24
25
26
27
28
29
30
0 0.85 1.15 1.35 1.55
Tem
per
atu
re (
Deg
ree
Cel
ciu
s)
Amplitude Current (mA)
Temperature versus Amplitude Current (Subject 3)
S1S2S3S4S5
mA
Univers
ity of
Mala
ya
53
Table 4.4: Temperature recorded of subject 4.
Current
(mA)
Time Temperature sensor (Degree Celcius)
S1 S2 S3 S4 S5
0 Initial 27.15 26.59 27.64 26.66 27.64
0.85 Minute 5 28.12 26.66 27.64 28.12 27.64
1.15 Minute 10 29.1 28.12 29.1 29.1 26.66
1.35 Minute 15 29.59 28.61 29.1 27.64 28.61
1.55 Minute 20 29.59 28.64 28.12 28.12 29.1
Figure 4.10 shows that the pattern of temperature increased of subject 4 was
different compared to subject 3. The graph shows that the performance of sensor to detect
temperature was not any better. The temperature detected was not constantly increased as
some of the sensors gave out a decreased of temperature during stimulation. But by the
end, the temperature still increased as predicted when the stimulation current increased. For
the initial temperature, value of S3 and S5 were the highest while value of S2 and S4 were
the lowest. Yet, at the end of stimulation, S1 gave out the highest value whereas S3 and S5
are the lowest.
Univers
ity of
Mala
ya
54
Figure 4.10: Graph of temperature versus amplitude current of subject 4.
Figure 4.11 shows the mean and standard deviation graph of skin temperature of
subject 3 and subject 4. From the graph, the pattern of temperature increased of subject 3
was better than subject 4. But, both shows an increased of mean temperature as the current
amplitude increased. Yet, the mean difference between end temperature and initial
temperature of subject 3 and subject 4 were different. For subject 3, the mean difference
was 1.09oC while for subject 4, the mean difference was 1.58
oC. It showed that during the
20 minutes stimulation, the rate of increased temperature for subject 4 was slightly higher
than subject 3. The standard deviation among the five sensors for both subject 3 and subject
4 was different for each value of current amplitude where the temperature was recorded.
Before stimulation, the standard deviation of temperature recoded was lower while when
the stimulation current was set at 1.15mA, the standard deviation of temperature between
the sensors was the highest.
25
26
27
28
29
30
0 0.85 1.15 1.35 1.55
Tem
per
atu
re (
Deg
ree
Cel
siu
s)
Amplitude Current (mA)
Temperature versus Amplitude Current (Subject 4)
S1
S2
S3
S4
S5
mA
Univers
ity of
Mala
ya
55
Figure 4.11: Graph of mean and standard deviation of skin temperature for subject 3 and
subject 4.
25
26
27
28
29
30
31
0 0,2 0,4 0,6 0,8 1 1,2 1,4 1,6Mea
n &
Sta
nd
ard
Dev
iati
on
(D
egre
ee C
elci
us)
Amplitude Current (mA)
Mean & Standard Deviation versus Amplitude Current
(Subject 3 & 4)
Subject 3
Subject 4
mA0.2 0.4 0.6 0.8 1 1.2 1.4 1.6
Univers
ity of
Mala
ya
56
CHAPTER 5: DISCUSSION
This project was to develop a heat feedback parameter of FES stimulator. Thus,
temperature sensor was used to detect the changes of skin temperature during the
stimulation muscles. One of the objectives was to examine the correlation between
activation of muscle and the heat produced by the muscles. This project had two parts
which were first part involved with the construction of FES stimulator and second part
involved with the testing of the temperature sensor.
For the first part, the FES stimulator was build to give an output of 80Hz, 32% of
duty cycle, and the current amplitude can be modulate to 40mA or more. After successfully
construct the stimulator circuit, the output was displayed and as predicted, it produced a
satisfying output. The current amplitude was set in range of 40mA to 70mA because this
project used only able-bodied subject. The values were half of the range of value used to
stimulate a SCI patient. Previous studies used a range of 120mA to 140mA current
amplitude to stimulate the muscles of SCI patients (Decker et al., 2010; Szecsi et al., 2008).
However, when the stimulator was connected to the stimulation electrode and tested
it to the subject, the subject did not feel any tingling sensation. This means that the current
output did not stimulate the subject’s muscles. After completing few modifications towards
the stimulator circuit, it still gave out a same result. It can be assume that there were few
Univers
ity of
Mala
ya
57
reasons that lead to the problem. First, maybe because of the amplitude voltage was not
high enough to activate the stimulation of pulse current. The voltage output for the
stimulator was only 9V compared to the FES stimulators within the market; they produced
more than 100V of voltage amplitude. Second, the stimulation electrode and electrode gel
have their own resistance that the current amplitude must pass through. Thus, it might have
reduced the current output. Third, maybe the current output could not go through the skin
impedance of human skin as the current output had been reduced. The resistance of human
skin was in range of 500Ω to 2kΩ depends on any part of the body. In the end, the
stimulator was successful to produce a satisfying output but still failed to stimulate the
muscles. It was concluded that the stimulator could not be used in Test 2 to stimulate the
muscle.
There were a few suggestions that might be able to overcome the problem faced by
the stimulator circuit. First, the output of stimulator can be connected to a transformer to
step-up the pulse voltage amplitude form 9V to over than 100V. This might produce a
strong stimulation pulse to stimulate the muscle. Second, if the output of transformer
produces low current amplitude, the value of resistor within the circuit can be change in
order to get higher current amplitude. By lowering the value resistor, it can increase the
amplitude of current output.
As the constructed stimulator has failed, it has been decided to replace the
stimulator with a previous FES stimulator that has been done by an undergraduate student.
However, the output of the previous stimulator was different than the constructed
stimulator. The pulse frequency produced was approximate to ≈17Hz, the amplitude
voltage was ≈7V and maximum current amplitude was only 1.55mA. But the stimulator did
Univers
ity of
Mala
ya
58
success to stimulate the muscles. Thus, Test 2 used the previous stimulator to stimulate the
muscles for 20 minutes.
For this project, only four men were used as the subjects. This is because it had been
mentioned in a previous study that women have a thicker layer of fat under their skin
compared to men. Although the skin thickness was not significantly different between men
and women, but women have a thicker subcutaneous fat in average of 0.7cm at the thigh
area while men only 0.51cm. It also showed that as the subcutaneous fat increased, the
stimulation current also needs to be increased (Petrofsky et al., 2008).
As seen the result in Chapter 4, the initial temperature of skin thigh for each subject
was different. The minimum temperature recorded was 24oC and the maximum value was
29oC. Previous study mentioned that the skin temperature for thigh was 31
oC. But there are
a few factors that lead to the change of skin temperature without any voluntary exercise.
The main factor was the temperature environment. A study proved that when a person
wears heavy clothing, the skin temperature increased because it received the inner body
temperature, whereas when the clothes were loosen, the skin temperature was decreased to
the environment temperature (Benedict et al., 1919). This explains the reason why the
temperature of thigh of the subjects displayed in range of temperature lower than 31oC.
Thus, the temperature sensor used in this project was not accurate enough to only
detect the true skin temperature. It was proposed in previous study, an apparatus which is
nearly instantaneous in action and sufficiently protected from the environment should be
used to record the true skin temperature, not the resultant of skin and environmental
temperature (Benedict et al., 1919). However, all five sensors used in this project indeed
success in detecting the increased of skin temperature for both Test 1 and Test 2.
Univers
ity of
Mala
ya
59
Test 1 and Test 2 were carried out in order to examine the correlation between
resistance of cycling, current amplitude and skin temperature. For Test 1, theoretically
when the resistance increased, subject must give more effort to maintain the performance of
the cycling, thus the activation of thigh muscles especially vastus lateralis was increased.
Thus, it expected that the heat produced by muscle will change the skin temperature. As
mentioned before, this test was voluntary experiment. As for Test 2, it involved non-
voluntary experiment. Theoretically, as the amplitude current increased, the stimulation
towards the muscles also increased and forced the muscles to activate more. Previously
mentioned for Test 1, as muscle activation increased, heat produced by muscle would
increase.
During the 20 minutes of experiment for both tests, it was expected that the
temperature increased as resistance and current amplitude increased. However, the mean
difference of temperature value between end value and initial value was different for both
Test 1 and Test2. The mean difference of skin temperature for Test 1 was between 2oC and
3oC while for Test 2, the mean difference was only between 1
oC to 1.59
oC. The only factor
that leads to the different of those values was because of the muscle activation. For Test 1,
as the resistance increased, the subject was voluntarily put an effort towards the muscles to
overcome the resistance, thus muscle activation increased. A study mentioned that the skin
temperature was affected by the skin blood flow. The temperature rises as the skin blood
flow increased (Petrofsky et al., 2008). Thus, it have been assumed that, when muscle
activation increased, blood would flow more to the muscles to transfer some oxygen from
the blood to the muscles, thus lead to the increased of skin temperature.
Univers
ity of
Mala
ya
60
For Test 2, the mean difference between initial temperature and final temperature
was lower because the current amplitude used was not strong enough to increase the
activation of muscles. During the stimulation, indeed the subjects feel the presence of pulse
current but as the current amplitude increased, the subjects only feel a little pain.
Supposedly, as current amplitude increased, subject would feel more pain. This concludes
that the maximum current of 1.55mA cannot be used in a test to examine the muscle
activation; it should be increased to 40mA or more as it had been proposed for able-bodied
person through a previous study (Decker et al, 2010). By doing this, in an increase the
muscle force thus will result an increase of muscle activation.
Univers
ity of
Mala
ya
61
CHAPTER 6: CONCLUSION
6.1 Summary
In summary, for the first part of this project, the objective of building a FES
stimulator has failed due to some reasons. First, due to the lower value of amplitude voltage
that was not enough to activate the stimulation of pulse current. Second, due to the
resistance of stimulation electrode and electrode gel, it might have reduced the stimulation
current. Third, the reduced stimulation current could not go through the skin impedance of
human skin. The stimulator indeed success produced a satisfying output with a pulse
frequency of 80Hz and current amplitude above 40mA, but unfortunately the current
produce did not reach to the subject’s muscles thus failed to stimulate the muscles.
Supposedly, the stimulator would be used for Test 2, but it was replaced by the previous
stimulator done by an undergraduate student.
Yet, this project successfully reached its objectives to detect the muscle contraction
level using heat as the feedback parameter and to examine the relationship between muscle
contraction level and heat production by the muscle. The hypothesis has been proven that
when the contraction level increased, the heat produced by muscle increased and the skin
temperature increased.
Univers
ity of
Mala
ya
62
Test 1 and Test 2 showed that there was a correlation between resistances of
cycling, current amplitude and the skin temperature. For Test 1, as resistance increased, the
skin temperature increased. For Test 2, as the current amplitude increased, the skin
temperature also increased. But the result performance of Test 1 was better than Test 2
because Test 2 was only used a small value of current amplitude. By the end, making the
muscle activation level increased, the increased of heat produced by stimulated muscles
was transferred to skin surface, thus also increased the skin temperature.
6.2 Recommendation for Future Work
For a future work, the stimulator that has failed will be modulated in order to
success produce pulse current and can stimulate the muscle. For examining the muscle
contraction level of other muscles, the output will be multiplex to stimulate other muscles
at the same time. Multiplexer will be used in advance.
Current study was using an open-looped system of FES stimulator; means that the
user need to manually modulate the current amplitude. For future work, a feedback circuit
with a microcontroller will be designed for a closed-looped system of FES stimulator. It
would remove the need for time uncontrollable manual adjustment. However, for this
application, it is important to calculate the expected accuracy of the detected sensor in order
to integrate this information into the parameter’s controller design. Thus, a more sensitive
temperature sensor will be used to read the true temperature of the skin surface.
Univers
ity of
Mala
ya
63
APPENDIX A
Arduino programming for TMP36:
//TMP36 Pin Variables
int sensorPin0 = 0; //the analog pin the TMP36's Vout (sense) pin is connected to
int sensorPin1 = 1; //the resolution is 10 mV / degree centigrade with a
int sensorPin2 = 2; //500 mV offset to allow for negative temperatures
int sensorPin3 = 3;
int sensorPin4 = 4;
/*
* setup() - this function runs once when you turn your Arduino on
* We initialize the serial connection with the computer
*/
void setup()
Serial.begin(9600); //Start the serial connection with the computer
//to view the result open the serial monitor
void loop() // run over and over again
//getting the voltage reading from the temperature sensor
int reading0 = analogRead(sensorPin0);
int reading1 = analogRead(sensorPin1);
int reading2 = analogRead(sensorPin2);
int reading3 = analogRead(sensorPin3);
int reading4 = analogRead(sensorPin4);
// converting that reading to voltage, for 3.3v arduino use 3.3
float voltage0 = reading0 * 5.0;
voltage0 /= 1024.0;
float voltage1 = reading1 * 5.0;
voltage1 /= 1024.0;
float voltage2 = reading2 * 5.0;
voltage2 /= 1024.0;
Univers
ity of
Mala
ya
64
float voltage3 = reading3 * 5.0;
voltage3 /= 1024.0;
float voltage4 = reading4 * 5.0;
voltage4 /= 1024.0;
// now print out the temperature
float temperatureC0 = (voltage0 - 0.5) * 100 ; //converting from 10 mv per degree wit 500
mV offset
float temperatureC1 = (voltage1 - 0.5) * 100 ;
float temperatureC2 = (voltage2 - 0.5) * 100 ;
float temperatureC3 = (voltage3 - 0.5) * 100 ;
float temperatureC4 = (voltage4 - 0.5) * 100 ; //to degrees ((volatge - 500mV) times 100)
Serial.print(temperatureC0); Serial.println(" degrees C0");
Serial.print(temperatureC1); Serial.println(" degrees C1");
Serial.print(temperatureC2); Serial.println(" degrees C2");
Serial.print(temperatureC3); Serial.println(" degrees C3");
Serial.print(temperatureC4); Serial.println(" degrees C4");
delay(3000); //waiting a second
Univers
ity of
Mala
ya
65
APPENDIX B
Datasheet of LM555 timer
Univers
ity of
Mala
ya
66
Univers
ity of
Mala
ya
67
Univers
ity of
Mala
ya
68
Univers
ity of
Mala
ya
69
Univers
ity of
Mala
ya
70
APPENDIX C
Datasheet of AD5220 Digital Potentiometer
Univers
ity of
Mala
ya
71
Univers
ity of
Mala
ya
72
Univers
ity of
Mala
ya
73
BIBLIOGRAPHY
Alon, G., Kantor, G., & Ho, H. S. (1994). Effects Of Electrode Size On Basic Excitory
Responses And On Selected Stimulus Parameters. Journal of Orthopaedic & Sports
Physical Therapy, 20(1), 29-35.
Amft, O., Junker, H., Lukowicz, P., Troester, G., & Schuster, C. (2006). Sensing muscle
activities with body-worn sensors. In BSN 2006: International Workshop on
Wearable and Implantable Body Sensor Networks, Proceedings (pp. 138-141).
Benedict, F. G., Miles, W. R., & Johnson, A. (1919). The Temperature of Human Skin.
PHYSIOLOGY: BENEDICT, MILES AND JOHNSON, 218 - 222.
Cheng, K. W. E., Lu, Y., Tong, K.-Y., Rad, A. B., Chow, D. H. K., Sutanto, D., et al.
(2004). Development of a Circuit for Functional Electrical Stimulation. IEEE
TRANSACTIONS ON NEURAL SYSTEMS AND REHABILITATION
ENGINEERING, 12(No.1), 43-47.
Chesler, N., & Durfee, W. (1997). Surface EMG as a Fatigue Indicator During FES-
induced Isometric Muscle Contractions. Journal of Electromyograph Kinesiol, 7(1),
27-37.
Currier, D. P., & Mann, R. (1983). Muscular Strength Development By Electrical-
stimulation in Healthy-Individuals. Physical Therapy, 63(6), 915-921.
Decker, M. J., Griffin, L., Abraham, L. D., & Brandt, L. (2010). Alternating stimulation of
synergistic muscles during functional electrical stimulation cycling improves
endurance in persons with spinal cord injury. Journal of Electromyography and
Kinesiology, 20, 1163–1169.
Ew, K. H., Wee, C. L., Zhang, D. G., Zhu, K. Y., & Zheng, H. (2005). Effects of Functional
Electrical Stimulation relating to Leg Movement. Paper presented at the
Engineering in Medicine and Biology 27th Annual Conference, Shanghai, China.
Forrester, B. J., & Petrofsky, J. S. (2004). Effect of Electrode Size, Shape, and Placement
During Electrical Stimulation. The Journal of Applied Research, 4(2), 346-354.
Graham, G. M., Thrasher, T. A., & Popovic, M. R. (2006). The Effect of Random
Modulation of Functional Electrical Stimulation Parameters on Muscle Fatigue.
IEEE Transactions On Neural Systems And Rehabilitation Engineering, 14(1), 38-
45.
Univers
ity of
Mala
ya
74
Grant, L. (1988). Functional Electrical Stimulation. IEE Review, 443-445.
How, T. K. (2011). Development Of A Single Channel Electrical Stimulation (FES)
Stimulator. University Of Malaya.
Kesar, T., Chou, L.-W., & Binder-Macleod, S. A. (2008). Effects of stimulation frequency
versus pulse duration modulation on muscle fatigue. Journal of Electromyography
and Kinesiology, 18, 662–671.
Lorussi, F., Scilingo, E. P., Tesconi, M., Tognetti, A., & De Rossi, D. (2005). Strain
sensing fabric for hand posture and gesture monitoring. Ieee Transactions on
Information Technology in Biomedicine, 9(3), 372-381.
Peckham, P. H. (1987). Functional electrical stimulation: current status and future prospects
of applications to the neuromuscular system in spinal cord injury. Paraplegia,
25(3), 279-288.
Petrofsky, J., Schwab, E., Lo, T., Cuneo, M., & Lawson, D. (2007). The thermal effect on
the blood flow response to electrical stimulation. Medical Science Monitor, 13(11),
CR498-CR504.
Petrofsky, J., Suh, H. J., Gunda, S., Prowse, M., & Batt, J. (2008). Interrelationships
between body fat and skin blood flow and the current required for electrical
stimulation of human muscle. Medical Engineering & Physics, 30(7), 931-936.
Popovic, M. R. (2006). Transcutaneous Electrical Stimulation Technology for Functional
Electrical Therapy Applications. Paper presented at the EMBS Annual International
Conference.
Reichel, M., Breyer, T., Mayr, W., & Rattay, T. (2002). Simulation of the three-
dimensional electrical field in the course of functional electrical stimulation.
Artificial Organs, 26(3), 252-255.
Sabut, S. K., Kumar, R., & Mahadevappa, M. (2010). Design of an insole embedded foot
pressure sensor controlled FES system for foot drop in stroke patients. Paper
presented at the 2010 International Conference on Systems in Medicine and
Biology.
Szecsi, J., Krewer, C., Muller, F., & b, A. S. (2008). Functional electrical stimulation
assisted cycling of patients with subacute stroke: Kinetic and kinematic analysis.
Clinical Biomechanics, 23, 1086–1094.
Tepavac, D., & Schwirtlich, L. (1997). Detection and Prediction of FES-induced Fatigue.
Journal of Electromyograph Kinesiol, 7(No.1), 39-50.
Theurel, J., Lepers, R., Pardon, L., & Maffiuletti, N. A. (2007). Differences in
cardiorespiratory and neuromuscular responses between voluntary and stimulated
contractions of the quadriceps femoris muscle. Respiratory Physiology &
Neurobiology, 157(2-3), 341-347.
Univers
ity of
Mala
ya
75
Valli, P., Boldrini, L., Bianchedi, D., Brizzi, G., & Miserocchi, G. (2002). Effects of low
intensity electrical stimulation on quadriceps muscle voluntary maximal strength.
Journal of Sports Medicine and Physical Fitness, 42(4), 425-430.
Vanderthommen, M., & Duchateau, J. (2007). Electrical stimulation as a modality to
improve performance of the neuromuscular system. Exercise and Sport Sciences
Reviews, 35(4), 180-185.
Williamson, R., & Andrews, B. J. (2000). Sensor systems for lower limb functional
electircal stimulation (FES) control. Medical Engineering & Physics, 22, 313–325.
Winslow, J., Jacobs, P. L., & Tepavac, D. (2003). Fatigue compensation during FES using
surface EMG. Journal of Electromyography and Kinesiology, 13, 555–568.
References from internet
http://www.diy-electronic-projects.com/p231-Muscular-Bio-Stimulator
http://www.ladyada.net/learn/sensors/tmp36.html
http://www.arduino.cc
http://www.fitstep.com/Advanced/Anatomy/Quadriceps.htm
http://www.shopcompex.com/training/electrode-placements/quadriceps
Univers
ity of
Mala
ya